Project 3 The principal goal of Project 3 is to better link the behaviors of zebrafish to underlying synaptic circuitry. In both the central and peripheral nervous systems we will leverage recent advances in high throughput serial section electron microscopy to acquire high resolution image data and then use new machine learning approaches to analyze this data. This effort will use three kinds of approaches: 1) sparse manual reconstructions of particular pathways related to particular behaviors, 2) dense reconstruction of synaptic connections to reveal all the input and output of neurons that is spatially correlated to markers that show excitatory and inhibitory cell types and activity, and 3) saturated reconstructions that also provide glial, vascular, sensory end organs and muscle (cardiac, skeletal and smooth). The zebrafish is the only vertebrate animal where obtaining a saturated reconstruction of the entire nervous system is conceivable. The connectomic data will be used initially to test the validity of published models of how a behavior is instantiated in the nervous system. In the likely event that there is some difference between what was expected based on published models and what is found, these data will be used to help formulate new models. Second, the data will be used to reveal previously unknown pathways, such as those that interconnect the autonomic and somatic nervous systems. Finally we will build the necessary infrastructure so the data can be used as a public, freely addressable atlas for any researcher who has interest in tracking the long and short range neural pathways that underlie vertebrate behavior at the resolution of single synapses.